Development of Autonomous Robot Osteotomy for Mandibular Ramal Bone Harvest and Evaluation of Its Accuracy: A Phantom Mandible-Based Trial

An autonomous robot osteotomy using direct coordinate determination for registering was developed, and the accuracy of the designed osteotomy along the preprogrammed plan was evaluated. Furthermore, the accuracy of the robotic and manual osteotomy was compared in regard to cut position, length, angle and depth. A light-weight robot was used in this study, with an electric gripper. Twenty stone models were used to evaluate accuracy of osteotomy and sixteen mandible phantoms were used to simulate the ramal bone harvest osteotomy for comparison between robotic and manual surgery. In the stone model experiment, the absolute mean values for osteotomy errors for position, length, angle, and depth were 0.93 ± 0.45 mm, 0.81 ± 0.34 mm, 1.26 ± 1.35°, and 1.19 ± 0.73 mm, respectively. In the mandible phantom model experiment, the robotic surgery showed lower errors for position, length and angle (0.70 ± 0.34 mm, 0.35 ± 0.19 mm and 1.32 ± 0.96°) and somewhat higher errors for depth (0.59 ± 0.46 mm) than manual surgery (1.83 ± 0.65 mm, 0.62 ± 0.37 mm, 5.96 ± 3.47° and 0.40 ± 0.31 mm). This study may provide a basis for developing clinical application of an autonomous robot osteotomy.